The continued increase in the number of vehicles in the transportation system calls for improving traffic safety and the efficiency of transportation infrastructure in general. With the advances in wireless communication network technology, the new emerging technology of what is known as Vehicle Dedicated Networks (VANET) has become an active area of research due to its high ability to improve road safety, road safety, and efficiency. In this paper, we present a methodology that manage traffics and reduce vehicle accidents and it consists of three stages, the first stage is collision detection according to vehicle position using k-means algorithm, the second stage is traffic management using SVM algorithm, and finally finding the optimal path for other vehicles to Crosses safely and smoothly through congestion area allowing emergency vehicles, ambulances, and firemen to get through if necessary. By using fuzzy logic.